After an impressive improvement observed in organic field effect transistors (OFETs) technology, there is a need to have a comprehensive model, which can predict output characteristics of OFETs to facilitate device-to-system integration in an efficient manner. In this paper, a global technique is developed to extract parameters of various OFETs models by evaluating their output characteristics. By involving respective model's expression, a simulator has been developed wherein the model variables have been extracted by using particle swarm optimization technique. Relative performance of a model is determined by assessing root mean square errors between the experimental and the modeled characteristics. Swarm optimization showed that the accuracy of a model is dependent upon the device dimensions. Thus, by employing the developed swarm optimization technique on devices of various dimensions, a preferred model is identified, which can be employed by the device modeling software. Swarm optimization showed that models, which are more physics based have a wider applicability than those which involved relatively higher number of fitting variables.